“If there are issues in the existing or legacy system, how would the “new” system being developed handle such issues? For instance, if the existing system had redundancy or it suffers from inconsistent data, will it be enough to make the data model and everything else will work fine?”
Migration of legacy systems is a good option when upgrading or integrating with a new system. In case of too many complications and increased costs, the old system needs to be discarded, but today with many best available practices and techniques for data migration, it should be possible to overcome this hurdle. Several of these best practices are iterative. Software development goes through various iterations, such as analysis, design, and testing. Similarly, deployment of the system follows an iterative cycle. When migrating it involves several cycles of rolling out the system to the users, coordinating old system with the new one. Multiple issues make data migration complex, and feature-rich data integration tools that support iterative migration are best suited for such scenarios. (Russom, 2006)
Russom (2006) states “In a recent TDWI Technology Survey, 41% of respondents chose ETL over hand-coding (27%), replication (11%), and EAI (3.5%). Users prefer ETL for its unique ability to handle the extreme requirements of data migration, including terabyte-scale datasets, multi-pass data transformations, deep data profiling, interoperability with data quality tools, and many-to-many data integration capabilities.”
Data migration is a process that should follow certain steps strictly to make it a success (White, 2014):
- It is essential to understand the data and the complexities attached to it. How it is currently stored and retrieved.
- There have to be data standards set in place to make successful use of data in the future.
- Understanding and outline the present and future business requirements.
- Defining of the roles and responsibilities of the people handling the data.
- Data migration does not mean just transferring or moving data from one point to another. Before doing so, the data should be examined, and any duplication should be eliminated. The data that is not used by the company should be removed.
- Focus on gathering the migration requirements.
- Assessing and selecting the proper tool for the new data environment.
- Data management automatically calls for risk management.
- Managing the change is essential. Everyone who uses the system & data should be on board & understand the new changes and data migration.
- Migration testing to ensure an error-free running system.
With technological advancements, it is vital to keep our systems up to date to make efficient use of them to enhance the business. Data migration is one of the best solutions for integrating to a new system.
References:
Russom, P. (2006) Best practices in data migration. Available at: http://download.101com.com/pub/tdwi/files/tdwi_monograph_bpindatamigration_april2006.pdf (Accessed: 23 January 2018)
White, W. (2014) 0 Data Migration Best Practices for any Organization. Available at: http://www.consultparagon.com/blog/10-data-migration-best-practices-for-any-organization (Accessed: 23 January 2018)
